CityCAT
CityCAT is a two-dimensional hydrodynamic urban flood model, developed for rapid assessment of pluvial and fluvial flood risk. It explicitly represents features including buildings, permeable surfaces and green infrastructure (Glenis, Kutija and Kilsby, 2018).
Pluvial and fluvial
Pluvial flooding is surface water: rain falling faster than the ground and drains can absorb it. Fluvial flooding is rivers overtopping their banks. Urban transport disruption is very often the first kind, which is why a model that routes water across streets and around buildings is the right tool here.
The word "explicitly" is doing real work in that description. Many flood models treat a city as a rough surface with an average roughness value. CityCAT puts the actual buildings into the computational domain, so water flows around them the way it does in reality, ponding where the geometry says it should.
Required input data
Four datasets define a CityCAT run. Three describe the ground; the fourth describes the weather hitting it.
Digital Terrain Model
| Source | Ordnance Survey (OS) Terrain 5 / EDINA Digimap |
| Resolution | 5 m |
| Description | A Digital Terrain Model representing ground surface elevation across the study area. The DTM provides the underlying topography for the CityCAT computational domain and determines the direction and routing of surface-water flow. |
This is the foundation. Water goes downhill, so the elevation model decides where it goes.
Green spaces
| Source | Ordnance Survey MasterMap Topography Layer / EDINA Digimap |
| Description | Spatial polygons representing permeable green areas within the study area. These data distinguish permeable from impermeable surfaces and represent rainfall infiltration within CityCAT. |
| Prepared as | GreenAreas.txt, the CityCAT-compatible format, with associated surface and soil parameters used to represent infiltration processes |
Buildings
| Source | Ordnance Survey MasterMap Topography Layer / EDINA Digimap |
| Description | Building footprint polygons representing the location and geometry of buildings within the study area. Incorporated explicitly into the computational domain to represent their influence on surface-water flow pathways and flood accumulation. |
| Prepared as | Buildings.txt, the CityCAT-compatible format |
Rainfall events
| Source | UKCP18 projections and rainfall scenarios developed for the DARe project |
| Description | Spatially and temporally varying rainfall events used as the meteorological forcing for the CityCAT simulations. Rainfall intensity varies through time and, where spatial rainfall data are available, across the modelling domain. |
Both current and future climate rainfall scenarios can be simulated, which is what makes it possible to assess how pluvial flood hazard changes under climate change.
Where the rainfall comes from
These events originate in the climate scenarios chapter. The weather generator produces the realisations; CityCAT consumes them.
Generated output data
CityCAT produces flood maps: spatially distributed flood-depth outputs representing the evolution of surface-water flooding across the study area.
Each output contains the simulated water depth for individual computational grid cells at the corresponding simulation time.
A time series, not a single worst case
This is the most consequential design decision in the chapter, and it is worth dwelling on.
In the DARe case studies, one flood-depth map is generated every 20 minutes throughout the simulation period. The result is a time series of flood conditions rather than only the maximum flood extent.
The sequential maps capture the full arc of the event: the development, peak and recession of flooding.
Why the maximum extent is not enough
A maximum-extent map tells you a road flooded at some point. It cannot tell you whether that happened during the morning peak or at three in the morning, or how long the road stayed closed.
Since the transport simulation downstream also runs in time, a time series lets the two be intersected properly. That is the difference between counting affected roads and counting affected journeys.
What happens to the output
These outputs can be processed in GIS and coupled with transport-network data to:
- Identify flooded road links
- Quantify the changing level of transport disruption over the course of the rainfall and flood event
That handover is covered in transport thresholds, which decides the depth at which a link is treated as disrupted, and in performance metrics, which turns the disruption into numbers.